How honest are the signals? A protocol for validating wearable sensors

Area of Science:

• Organizational behavior
• Human-computer interaction
• Data science

Background:

• Growing interest in alternative data sources beyond self-reports for measuring organizational behavior.
• Wearable sensors offer new avenues for data collection but their data veracity is largely unverified.
• Need for rigorous validation of wearable sensor data for theoretical development and practical application.

Purpose of the Study:

• To test the validity and reliability of data captured by a specific wearable badge (Humanyze) during structured meetings.
• To provide evidence-based recommendations for the appropriate use of wearable badge data in organizational research.
• To introduce a website enabling researchers to validate their own wearable badge data.

Main Methods:

• Series of studies targeting specific sensors of the Humanyze badge relevant to structured meetings.
• Controlled experimental design with participants wearing badges throughout meetings.
• Data analysis focused on sensor accuracy, consistency, and correlation with known meeting dynamics.

Main Results:

• Specific sensors on the wearable badge demonstrated acceptable validity and reliability for capturing meeting-related behaviors.
• Performance varied across different sensors, necessitating careful selection and interpretation of data.
• Validation results informed specific usage recommendations for researchers.

Conclusions:

• Wearable sensor data holds potential for organizational research, but rigorous validation is crucial.
• The study provides a framework and tools for assessing the quality of wearable sensor data.
• Researchers can leverage validated sensor data to enhance the understanding of organizational behavior and practice.